Max Planck Institute for Dynamics and Self-Organization -- Department for Nonlinear Dynamics and Network Dynamics Group
Personal tools
Log in


Tuesday, 10.12.2013 17:15 s.t.


by Dr. Anna Levina
from Max-Planck-Institut für Dynamik und Selbstorganisation, Göttingen

Contact person: Theo Geisel


Ludwig Prandtl lecture hall


Neuronal avalanches were recorded in many experiments on different systems and in different conditions. They have been shown to play important role in the information processing in the brain. In my talk I will discuss how learning and structural plasticity affect criticality in the neuronal network. Memory building Hebbian learning imprints structure from the external activation into the network and makes spontaneous re-occurrence of specific patterns more probable. Thus it has tendency to counteract criticality. I present the first example of the network that maintains critical state while keeping memory of the stored patterns. Not only imprinting of the memories can interact with criticality. Long term synaptic plasticity is known to have crucial influence on the functioning of the brain. I show that interplay of different plasticity types does not preclude criticality, moreover it makes network more stable in the critical state. In particular I will speak about spike timing dependent plasticity and homeostatic plasticity. Maximization of dynamic range is probably the most established benefit of critical network, that was also found experimentally. However only very schematic model investigated how the dynamical range is connected to the avalanche size distribution. Here I'm going to present some results from the more realistic network.

back to overview